|Publication number||US7085315 B1|
|Application number||US 10/122,784|
|Publication date||Aug 1, 2006|
|Filing date||Apr 15, 2002|
|Priority date||Apr 15, 2002|
|Publication number||10122784, 122784, US 7085315 B1, US 7085315B1, US-B1-7085315, US7085315 B1, US7085315B1|
|Inventors||James R Kelton|
|Original Assignee||Vixs, Inc.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (5), Referenced by (20), Classifications (7), Legal Events (6)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This invention relates generally to wireless communication systems and more particularly to receiving transmission within such wireless communication systems.
Wireless communication systems are known to include a plurality of wireless communication devices that communicate directly (e.g., point-to-point) or through an infrastructure. For direct communications, a wireless communication device, such as a radio, cellular telephone, station coupled to a personal computer or laptop, et cetera, transmits data on a particular radio frequency channel directly to another wireless communication device. For infrastructure-supported communications, a wireless communication device transmits data on an assigned radio frequency channel to an access point (or a base station). The access point determines the targeted wireless communication device from the received RF signals. If the targeted wireless communication device is affiliated with the access point, the access point transmits the data to the targeted wireless communication device on a radio frequency channel. If the targeted wireless communication device is not affiliated with the access point, the access point forwards the data to a central station, which routes the data to the access point that is affiliated with the targeted wireless communication device.
To ensure reliability of data transmissions within a wireless communication system and to ensure interoperability of differing manufacturers' equipment, standards have been developed. Such wireless communications standards include IEEE8 02.11, Bluetooth, advanced mobile phone services (AMPS), digital AMPS, global system for mobile communications (GSM), code division multiple access (CDMA), local multi-point distribution services (LMDS), multi-channel, multi-point distribution systems (MMDS), and/or variations thereof.
Such standards prescribe operating parameters for particular types of wireless communication systems. For example, the IEEE 802.11a standard defines a wireless local area network that prescribes a frequency band of use, division of the frequency band into channels and sub-channels, encoding/decoding convention, modulation/demodulation convention, frame format, data rates, et cetera. Further, the IEEE 802.11a standard provides various combinations of data rates and modulation schemes, which can be selected via a coding rate corresponding to a particular modulation scheme.
Within any standard compliant wireless communication system, a wireless communication device transmits and receives data via RF (radio frequency) communication paths, i.e. RF channels. To transmit data, a wireless communication device encodes and/or modulates the data in accordance with the applicable standard to produce encoded data. The wireless communication device then mixes the encoded data with a local oscillation in one or more stages to produce an RF signal. The wireless communication device transmits the RF signal via an antenna.
To receive data, a wireless communication device receives an RF signal via an antenna, filters the RF signal and amplifies it. The wireless communication device then mixes the filtered and amplified RF signal with a local oscillation in one or more stages to produce a baseband signal. In an ideal environment, the baseband signal will be identical to the encoded data produced by the wireless communication device that transmitted the RF signal. Since wireless communication devices do not operate in an ideal environment, the baseband signal produced by the receiver is not identical to the encoded data transmitted by the transmitter due to the transfer function characteristics of the RF channel and of the receiver. In particular, the baseband signal may be phase, amplitude, and/or frequency offset from the encoded signal.
To compensate for the phase, amplitude, and/or frequency offset, most receivers include frequency equalization (FEQ). One method of frequency equalization is to force frequency domain errors to zero. While this is the simplest approach, it forces some frequency components to have zero error that do not need correction, thus, in some instances, the zero forcing over compensates the baseband signal. Other methods include linear prediction and adaptive frequency equalization. Linear prediction calculates the frequency response of the RF channel to determine the appropriate response for frequency equalization. While more accurate than zero forcing, it is very complex thus expensive in cost and in integrated circuit real estate. Adaptive FEQ adjusts the frequency equalization at the data rate. For IEEE802.11a applications, adaptive frequency equalization is impractical for moderate to low cost solutions due to its complexity and speed requirements.
Therefore, a need exists for a method and apparatus that enables wireless communication devices to adequately and cost effectively compensate for the frequency transfer functions of the RF channel, the transmitter, and/or the receiver that adversely filters received signals.
Generally, the present invention provides a method and apparatus for digital demodulation for use in a wireless communication device. Such a digital demodulator includes processing that begins by performing a fast Fourier transform to convert a plurality of time domain symbol components representing inbound data into a plurality of frequency domain symbol components. The processing continues by frequency domain equalizing at least some of the plurality of frequency domain symbol components based on a dynamic compensation control signal to dynamically compensate for adverse transmission characteristics of the plurality of time domain symbol components. The processing then continues by generating a plurality of estimated symbols and a plurality of error values for the plurality of compensated frequency domain symbol components. The processing then continues by generating the dynamic compensation control signal based on at least one of the plurality of error values, the plurality of frequency domain symbol components and the plurality of compensated frequency domain symbol components. With such a method and apparatus, efficient and cost effective frequency equalization may be implemented in digital demodulators, which enable more robust wireless communication devices to be economically manufactured.
The present invention can be more fully described with reference to
In operation, the transmitter section 14 receives outbound data 32 via the encoding module 24 of the modulation module 16. The outbound data 32 may be digitized voice transmissions, data transmissions, and/or any other type of information that may be communicated over a wireless communication channel. The encoding module 24 encodes the outbound data 32 in accordance with a particular wireless communication standard. For instance, for the IEEE 802.11a standard, the encoding includes scrambling, convolutional encoding, and data compression. The interleaving and mapping module 26 receives the encoded data and interleaves and maps the data to particular constellation symbols: one symbol per sub-carrier of a channel. Each of symbols includes an I and a Q component. The inverse fast Fourier transform module 28 converts the I and Q components for each of the symbols from the frequency domain to the time domain to produce a plurality of time domain magnitudes and phases, i.e., time domain symbol components.
The compulator 30 receives the plurality of time domain symbol components and converts them into a baseband analog signal that is provided to the up conversion module 18. For example, if the wireless communication device 10 is compliant with the IEEE 802.11a standard, the inverse fast Fourier transform module is converting sixty-four I and Q components, which represent sixty-four symbols, into sixty-four time domain symbol components. Of the sixty-four symbols, only fifty-two are used of which forty-eight carry data and for are pilot signals. The compulator 30 combines the sixty-four time domain symbol components (i.e., magnitudes and phases) into the baseband analog signal that is provided to the up conversion module 18.
The up conversion module 18 converts the baseband analog signal received from the compulator 30 into a radio frequency (RF) signal by mixing the baseband analog signal with one or more local oscillations in one or more stages. Bandpass filter 20 filters the radio frequency signal and provides the filtered RF signal to the power amplifier 22. The power amplifier 22 increases the energy of the RF signal such that it may be transmitted as an outbound RF signal 34 via the antenna.
The receiver section 12 includes a low noise amplifier 36, which is coupled to receive inbound RF signals 56 via an antenna, a bandpass filer 38, a down conversion module 40, and a digital demodulator 42. The digital demodulator 42 includes a decompulator 44, a fast Fourier transform module 46, a frequency domain equalizer module 48, a slicer 50, a deinterleaving and demapping module 52, a decoding module 54 and a fixed filter feedback module 65.
The low noise amplifier 36, bandpass filter 38 and down conversion module 40 are operably coupled to receive inbound RF signals 56 and produce baseband signals corresponding thereto. The decompulator 44 converts the analog baseband signals into a plurality of digital time domain symbol components (i.e., magnitudes and phases) that correspond to the plurality of time domain symbol components transmitted by a transmitting wireless communication device that transmitted the inbound RF signal 56. As previously stated, the plurality of time domain symbol components represents a plurality of symbols, which, in turn, represents encoded data.
The fast Fourier transform module 46 converts the plurality of digital time domain magnitudes and phases, i.e., time domain symbol components, into a plurality of frequency domain symbol components 58, i.e., I components and Q components. For example, in an IEEE 802.11a compliant receiver, the plurality of frequency domain symbol components includes sixty-four symbol components (e.g., sixty-four I and Q components for sixty-four symbols). Of the sixty-four symbol components, twelve represent no data, forty-eight represent data, and four represent pilot data.
The frequency domain equalizing module 48 receives the plurality of frequency domain symbol components 58 and individually compensates them for droop caused by the receiver section 12 and also compensates them for phase, frequency, and/or amplitude offsets caused by the transmission characteristics of the RF channel based on a dynamic compensation control signal 66, which includes filtering coefficients for each bin. The fixed filter feedback module 65 generates the dynamic compensation control signal 66 based on either the compensated frequency domain symbol components 60 or the estimated symbol values 62 and error values 64. While the frequency domain equalizing module 48 individually compensates each frequency domain symbol component to produce a compensated frequency domain symbol component 60, the fixed filter feedback module 65 estimates the filtering coefficients for each bin, and, for most bins, estimates the filtering coefficients based on the filtering coefficients of adjacent bins. As one of average skill in the art will appreciate, a bin contains one symbol (or symbol components) and when modulated is transmitted via one sub-carrier of a channel. The frequency domain equalizing module 48 and the fixed filter feedback module 65 will be described in greater detail with reference to
The slicer 50 receives the compensated frequency domain symbol components 60 and produces an estimated symbol value 62 and an error value 64 for each symbol component 60. In general, the slicer 50 maps each actual symbol 60 to the nearest valid constellation point (i.e., the estimated symbol value) and determines a difference vector between the actual symbol and the nearest constellation point (i.e., estimated error). The estimated symbol value 62 and corresponding error value 64 are more fully illustrated in
Returning to the discussion of
As the transmitter section 14 of radio #1 is preparing the RF signals 56 for transmission, the transmitter section 14 is introducing phase, frequency, and/or amplitude offsets due to its transfer function [Tx(ω)]. As the RF signals traverse the RF channel 70 additional phase, frequency, and/or amplitude offset is introduced due to the transfer function of the RF channel [G(ω)]. As such, by the time the RF signals reach the antenna of the receiver section 12 of radio #2 a significant amount of phase, frequency, and/or amplitude offset has been introduced. Further phase, frequency, and/or amplitude offset is introduced by the low noise amplifier 36, the bandpass filter 38, and the down, which, collectively, have a transfer function [Rx(ω)].
The phase, frequency, and/or amplitude offset is propagated through the decompulator 44 and the FFT module 46 such that each of the plurality of frequency domain symbol components 58 include a phase, frequency, and/or amplitude offset. While the phase, frequency, and/or amplitude offset will vary from frequency domain symbol to frequency domain symbol, the phase, frequency, and/or amplitude offset of adjacent frequency domain symbols will be have a strong correlation. The function of the FEQ module 48 is to substantially undo the phase, frequency, and/or amplitude offset introduced by the transmitter, the RF channel, and the receiver.
To remove the phase, frequency, and/or amplitude offset introduced by the receiver, the 1st FEQ module 72 includes a plurality of single tap filters 76–78 that collectively have a compensation characteristic of 1/Rx(ω). Each single tap filter 76–78 filters a bin (i.e., a frequency domain symbol component) of a channel (e.g., an IEEE 802.11a compliant channel includes 64 bins) to filter the corresponding filtering introduced by the receiver section. The filtering performed by the 1st FEQ module 72 can be more fully illustrated with reference to
Since the receiver transfer function is fixed due to the circuitry embodying the radio receiver, the coefficients for the single tap filters of the 1st FEQ may be readily determined and do not vary significantly over operating conditions. As such, the coefficients for the 1st FEQ may be fixed to provide the desired frequency compensation for the droop caused by the radio receiver.
Returning to the discussion of
Returning to the discussion of
To begin, the coefficients are initialized during a training sequence during which time known signals of known phase, frequency, and amplitude are transmitted. Since the phase, frequency, and amplitude of the training signals are known, the coefficients of the programmable filters 80–82 are established to compensate for the difference in phase, frequency, and/or amplitude of the signals that were received and the known phase, frequency, and amplitude of the training signals.
Having initialized the coefficients, the fixed filter feedback module 65 dynamically updates them as needed. This is accomplished by determining weighted offsets 92 by the temporal and frequency weighting module 74 from the error values 64. In general, the temporal and frequency weighting module 74 is determining a difference between the current coefficient for one of the programmable filters 80–82 and what it should be to have an error value of zero. But, since the dominate source for error is noise once the coefficients are close to an optimum value, the correction determined by the temporal and frequency weighting module 74 is weighted, thus producing the weighted offset 92. In one embodiment, the temporal and frequency weighting module 84 may perform a least mean square function to estimate the error between the present coefficient and the present channel response and generate the weighted offset such that the coefficients are adjusted in a direction to reduce the error.
The smoothing filter 86, which includes the N-tap filter 88 and the interpreting module 90, receives the weighted offsets 92 and generates the dynamic compensation control signals 66 therefrom. As previously mentioned, the dynamic compensation signal 66 includes the coefficients for each of the programmable filters 80–82. As one of average skill in the art will appreciate, the dynamic compensation signal 66 may continually provide the coefficients for each programmable filter regardless of whether the coefficients change, or the dynamic compensation signal 66 may only provide coefficients to certain programmable filters when the coefficients change.
To produce the dynamic compensation control signal 66, i.e., coefficients for the plurality of programmable filters 80–82, the N-tap filter 88 filters the weighted offsets 82 in accordance with an interpretation 94 provided by the interpreting module 90. In general, the N-tap filter utilizes the coefficient values of adjacent bins (i.e., adjacent programmable filters that filter adjacent frequency domain symbol components) to determine the coefficients for the current bin by performing an N-tap filtering function. The interpreting module 90 monitors the adjacent bins to determine if they are empty (i.e., carrying no data). If so, the interpreting module 90 provides the interpretation 94 that causes N-tap filter 88 to alter its filtering properties for the current bin.
The N-tap filter 88 performs a filtering function of
where ia=(NJ−(N−1)/2) and N is an odd number equal to or greater than three and Kj are selectable constants. For example, if N equals three, then the filtering function is equal to ⅛H(i−1)+¾Hi+⅛H(i+1). As a further example, if N equals five, then the filtering function is equal to 1/16H(i−2)+⅛H(i−1)+⅝Hi+⅛H(i+1)+ 1/16H(i+2). The N-tap filter 88 performs this filtering function unless one of the adjacent bins in zero.
In the case where one of the adjacent bins is zero, the interpreting module 90 generates the interpretation 94, which causes the N-tap filter to alter the filtering function to HiF=Hi and
which performs the similar function as above, but with the bin containing zero removed from the filtering function. For example, if the second alternative is used, the filtering function may be may be HiF=⅞Hi+⅛H(i+1) if the preceding sub-carrier is containing zero information or HiF=⅞Hi+⅛H(i−1) if the subsequent sub-carrier is carrying zero data.
The process then proceeds to Step 124 where the digital demodulator generates a plurality of estimated symbol values and a plurality of error values from the plurality of compensated frequency domain symbol components. The process then proceeds to Step 126 where the digital demodulator determines the dynamic compensation control signal based on the plurality of error values and the plurality of frequency domain symbol components or based on the plurality of compensated frequency domain symbol components. As such, the digital demodulator is determining the coefficients for each of a plurality of programmable filters contained within the frequency equalizer, which substantially filters out the adverse effects introduced by the RF channel.
The preceding discussion has presented a method and apparatus for digital demodulation, which includes frequency equalization. As one of average skill in the art will appreciate, other embodiments may be derived from the teaching of the present invention, without deviating from the scope of the claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US6314147 *||Nov 4, 1998||Nov 6, 2001||The Board Of Trustees Of The Leland Stanford Junior University||Two-stage CCI/ISI reduction with space-time processing in TDMA cellular networks|
|US6549512 *||Jun 25, 1998||Apr 15, 2003||Texas Instruments Incorporated||MDSL DMT architecture|
|US6628730 *||Jan 16, 2001||Sep 30, 2003||Lsi Logic Corporation||Demodulating digital video broadcast signals|
|US6735255 *||Feb 29, 2000||May 11, 2004||3Com Corporation||Correlation based method of determining frame boundaries of data frames that are periodically extended|
|US6834079 *||Oct 20, 2000||Dec 21, 2004||3Com Corporation||Efficient implementation for equalization of multicarrier channels|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7324608 *||Feb 18, 2004||Jan 29, 2008||Ralink Technology, Inc.||Efficient subcarrier weighting to enhance receiver performance|
|US7356108 *||Apr 6, 2004||Apr 8, 2008||Advanced Micro Devices, Inc.||OFDM receiver having adaptive channel estimator for correcting channel fading based on accumulated pseudo power values|
|US7787557 *||Feb 27, 2007||Aug 31, 2010||Samsung Electronics Co., Ltd.||Apparatus for equalizing clipping noise signals of receiver systems and method thereof|
|US7813439||Feb 6, 2006||Oct 12, 2010||Broadcom Corporation||Various methods and apparatuses for impulse noise detection|
|US7852950||Feb 25, 2005||Dec 14, 2010||Broadcom Corporation||Methods and apparatuses for canceling correlated noise in a multi-carrier communication system|
|US7953163 *||May 16, 2005||May 31, 2011||Broadcom Corporation||Block linear equalization in a multicarrier communication system|
|US8194722||Oct 11, 2005||Jun 5, 2012||Broadcom Corporation||Various methods and apparatuses for impulse noise mitigation|
|US8472533||May 2, 2011||Jun 25, 2013||Broadcom Corporation||Reduced-complexity common-mode noise cancellation system for DSL|
|US8571156||May 27, 2008||Oct 29, 2013||Nxp, B.V.||Receiver apparatus|
|US8605837||Oct 10, 2008||Dec 10, 2013||Broadcom Corporation||Adaptive frequency-domain reference noise canceller for multicarrier communications systems|
|US8995502 *||Mar 20, 2007||Mar 31, 2015||Apple Inc.||Transceiver with spectral analysis|
|US9014318 *||Mar 25, 2013||Apr 21, 2015||Micron Technology, Inc.||Self-calibrating continuous-time equalization|
|US9160381||Dec 6, 2013||Oct 13, 2015||Broadcom Corporation||Adaptive frequency-domain reference noise canceller for multicarrier communications systems|
|US9374257||Mar 15, 2006||Jun 21, 2016||Broadcom Corporation||Methods and apparatuses of measuring impulse noise parameters in multi-carrier communication systems|
|US20080089224 *||Feb 27, 2007||Apr 17, 2008||Samsung Electronics Co., Ltd.||Apparatus for equalizing clipping noise signals of receiver systems and method thereof|
|US20100303161 *||May 27, 2008||Dec 2, 2010||Nxp B.V.||Receiver apparatus|
|US20130230089 *||Mar 25, 2013||Sep 5, 2013||Micron Technology, Inc.||Self-calibrating continuous-time equalization|
|CN101682601B||May 27, 2008||Oct 3, 2012||Nxp股份有限公司||Frequency domain interpolation based receiver for multicarrier signals|
|WO2008146242A2 *||May 27, 2008||Dec 4, 2008||Nxp B.V.||Frequency domain interpolation based receiver for multicarrier signals|
|WO2008146242A3 *||May 27, 2008||May 28, 2009||Nxp Bv||Frequency domain interpolation based receiver for multicarrier signals|
|U.S. Classification||375/219, 375/229|
|Cooperative Classification||H04L2025/03414, H04L25/03159, H04L2025/03522|
|Apr 15, 2002||AS||Assignment|
Owner name: VIXS, INC., ONTARIO
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:KELTON, JAMES R.;REEL/FRAME:012809/0947
Effective date: 20020403
|Feb 9, 2007||AS||Assignment|
Owner name: VIXS SYSTEMS, INC., ONTARIO
Free format text: CORRECTIVE ASSIGNMENT TO CORRECT THE CHANGE ASSIGNEE NAME PREVIOUSLY RECORDED ON REEL 012809 FRAME 0947;ASSIGNOR:KELTON, JAMES R.;REEL/FRAME:018877/0954
Effective date: 20020430
|Feb 11, 2009||AS||Assignment|
Owner name: COMERICA BANK, CANADA
Free format text: SECURITY AGREEMENT;ASSIGNOR:VIXS SYSTEMS INC.;REEL/FRAME:022240/0446
Effective date: 20081114
Owner name: COMERICA BANK,CANADA
Free format text: SECURITY AGREEMENT;ASSIGNOR:VIXS SYSTEMS INC.;REEL/FRAME:022240/0446
Effective date: 20081114
|Jan 20, 2010||FPAY||Fee payment|
Year of fee payment: 4
|Jan 2, 2014||FPAY||Fee payment|
Year of fee payment: 8
|Aug 18, 2017||AS||Assignment|
Owner name: VIXS SYSTEMS, INC., CANADA
Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:COMERICA BANK;REEL/FRAME:043601/0817
Effective date: 20170802